[Freesurfer] Troubleshooting: Soft Failures

External Email - Use Caution

Hi FreeSurfer experts,

Recently, I used Fresurfer's recon-all toprocess our own T1 MRI brain volumes which is 0.7 m3 isotropic. For most young brain cases, FS performs very well, the segmention is great. But for almost every elderly brain case, the segmention is inaccurate visibly. The figure is the same slice of T1 data and aseg. mgz , the GM is mislabeled as wm-hypointensities.

I found some solutions to this problem. But these solutions all require manual editing, which will take a long time. Is there a way to automatically solve this problem by setting some parameters?

Thank you for your attention and your help!

Cui